Alarm "High pNic error rate detected" on vSAN cluster

Products

• VMware vSAN

Issue/Introduction

In vSAN environments, especially those integrated with VMware NSX, you may encounter the alarm "High pNic error rate detected" on a specific physical network interface card (pNIC). This alarm is a key indicator of potential packet loss, which can degrade vSAN performance and stability.
 · Alarm triggered for a vSAN uplink (e.g., vmnic#).
· Running the command esxcli network nic stats get -n vmnic#  reveals a high and incrementing counter for Rx Length Errors
· The environment has VMware NSX deployed.

Environment

vSAN 

Cause

The "Rx Length Errors" counter increments when a NIC receives a data frame that exceeds its configured Maximum Transmission Unit (MTU) size. The frame is then dropped by the NIC hardware. In a vSAN+NSX environment, this is almost always caused by an MTU size mismatch between the virtual/overlay network and the physical underlay network.

Resolution

If certain private network protocols have already been excluded, such as Cisco's WLCCP, Cisco DTP, and some STP packets, these packets inherently have this special frame structure, which can also cause RX length errors. However, these packets will not damage or affect the performance of the physical network and are harmless.
It has been ensured that the external network is not causing these RX length errors by sending a large number of multicast and broadcast packets.

To set ESXI physical NIC MTU to be greater than 9100.
Please Note: Ensure physical Switch Port must be equal to or greater than the MTU set on the ESXi physical NIC.